Struct DataHeuristics 

pub struct DataHeuristics {
    pub block_size: usize,
    pub row_alignment: RowAlignment,
    pub min_repeat_rows: Option<usize>,
    pub interior: Vec<RunRule>,
    pub leading: Option<EdgeTrim>,
    pub trailing: Option<EdgeTrim>,
    pub all_zero_single_ds_min: usize,
    pub min_aligned_blocks: Option<usize>,
    pub zero_skip_windowed: bool,
}

Fields

block_size: usize

Bytes per .db row; also the unit for alignment and row-repeat.

row_alignment: RowAlignment

How literal rows break for display; does not affect byte layout.

min_repeat_rows: Option<usize>

Collapse N+ identical block_size rows into one .rept (hexdump *). None disables multi-byte row folding.

interior: Vec<RunRule>

RLE rules applied to runs in the body of the span.

leading: Option<EdgeTrim>

Trim at the start of a mapped chunk (None = leave leading bytes literal).

trailing: Option<EdgeTrim>

Trim at the end of a mapped chunk (None = leave trailing bytes literal).

all_zero_single_ds_min: usize

When the entire span is zeros and length is at least this value, emit one skip run for the whole span. usize::MAX disables the fast path.

min_aligned_blocks: Option<usize>

When set, run compression only applies to the address-aligned middle of a homogeneous run spanning at least this many block_size blocks.

zero_skip_windowed: bool

Emit zero skip runs as multiple .ds block_size lines instead of one .ds N for the full run length.

Implementations

impl DataHeuristics

pub fn iterate<'a>( &self, address: AddressValue, edge: Option<Edge>, bytes: &'a [u8], ) -> impl Iterator<Item = DataChunk<'a, u8>>

pub fn literal_rows<'a>( &self, address: AddressValue, bytes: &'a [u8], ) -> Vec<&'a [u8]>

Split a literal span into .db rows per block_size and row_alignment.

Trait Implementations

impl Default for DataHeuristics

fn default() -> Self

Returns the “default value” for a type.